Gastrointestinal stromal tumors (GISTs), the most common type of adult soft tissue sarcoma, are diagnosed in approximately 5,000-10,000 Americans each year. Genomic activation of KIT or PDGFRA receptor tyrosine kinases plays a role in the development of ~85% of GISTs. Imatinib, a KIT/PDGFRA tyrosine kinase inhibitor (TKI), has revolutionized the treatment of patients with advanced, unresectable GIST. Nevertheless, primary resistance to the drug is observed in some patients and secondary (acquired) resistance occurs with increasing frequency over time. The median time to tumor progression during front-line imatinib therapy is only 18-22 months. Notably, in the majority of treated GISTs, imatinib resistance results from selection of clones with drug-resistant kinase mutations. Hypothesis: the most efficient pathway for optimizing the treatment of patients with advanced GIST is to identify the relevant mechanisms of TKI-resistance and establish model systems that allow the testing of alternative or complementary agents, including combination therapy. Specific Aim 1. Using cell-based models, we will identify and validate treatment strategies to circumvent and/or prevent the emergence of secondary resistance to TKIs caused by the selection of tumor subclones with acquired KIT kinase mutations. This aim will utilize two different cellular models: 1) a transient transfection expression system to biochemically profile the activity of our panel of TKIs against previously identified imatinb-resistant mutant kinases;and 2) an ENU-mutagenesis/TKI drug selection system to identify novel KIT kinase mutations that confer TKI-resistance. Specific Aim 2. To determine the frequency and spectrum of secondary KIT (or PDGFRA) kinase mutations present in TKI-resistant clinical specimens. Clinical specimens will be obtained from patients treated with two or more successive TKIs (e.g imatinib followed by sunitinib). Novel kinase mutations that are identified will be cloned and biochemically characterized as described above. Potential impact on Veterans Health Care: Cancer is a major cause of morbidity and mortality in American veterans, especially in veterans over the age of 50. The development of new cancer therapeutic agents that target pathogenetic or significant aspects of tumor biology would greatly increase the quality and duration of life for veterans with cancer. PUBLIC HEALTH RELEVANCE: Relevance to Veterans Health Cancer is a major cause of morbidity and mortality in American veterans, especially in veterans over the age of 50. The care of veterans with cancer represents a significant portion of the overall Veterans Affairs Health Care budget. The development of new cancer therapeutic agents that target pathogenetic or significant aspects of tumor biology will greatly increase the quality of life of veterans with cancer. Such targeted agents have the potential to be much more effective than conventional treatments and with significantly less toxicity. Reducing toxicity of cancer treatments should reduce cancer care costs by decreasing hospitalizations for side effects of cancer treatment (such as neutropenic fever and/or mucositis). Decreasing treatment-related medical complications is of particular relevance to the treatment of veterans who often have significant co-morbidities such as cardiopulmonary disease and/or diabetes that increase the risk of adverse reactions to conventional cytotoxic agents. The model to be studied in the current application has direct relevance to patients with Gastrointestinal Stromal Tumors, but will likely provide major insights into cancer biology and the mechanisms of resistance to other small molecule kinase inhibitors. These insights will be directly applicable to improving the treatment of other malignancies that are dependent upon KIT and/or PDGFRA signaling (e.g. melanoma, acute myeloid leukemia, seminoma, and mastocytosis). However, the results for this project will also be valuable for the evolving development of molecularly targeted therapies for the more common solid tumors such as lung, colon, prostate, and breast cancer.